Huang et al. Ann.

Intensive Care (2020) 10:49

https://doi.org/10.1186/s13613-020-00662-y

REVIEW Open Access

Systematic review and literature appraisal

on methodology of conducting and reporting
critical‑care echocardiography studies: a report
from the European Society of Intensive Care
Medicine PRICES expert panel
S. Huang1, F. Sanfilippo2, A. Herpain3, M. Balik4, M. Chew5, F. Clau‑Terré6, C. Corredor7, D. De Backer8,
N. Fletcher9, G. Geri10,11, A. Mekontso‑Dessap12, A. McLean1, A. Morelli13, S. Orde1, T. Petrinic14, M. Slama15,
I. C. C. van der Horst16, P. Vignon17, P. Mayo18 and A. Vieillard‑Baron10,11*

Abstract 
Background:  The echocardiography working group of the European Society of Intensive Care Medicine recognized
the need to provide structured guidance for future CCE research methodology and reporting based on a systematic
appraisal of the current literature. Here is reported this systematic appraisal.
Methods:  We conducted a systematic review, registered on the Prospero database. A total of 43 items of common
interest to all echocardiography studies were initially listed by the experts, and other “topic-specific” items were sepa‑
rated into five main categories of interest (left ventricular systolic function, LVSF n = 15, right ventricular function, RVF
n = 18, left ventricular diastolic function, LVDF n = 15, fluid management, FM n = 7, and advanced echocardiography
techniques, AET n = 17). We evaluated the percentage of items reported per study and the fraction of studies report‑
ing a single item.
Results:  From January 2000 till December 2017 a total of 209 articles were included after systematic search and
screening, 97 for LVSF, 48 for RVF, 51 for LVDF, 36 for FM and 24 for AET. Shock and ARDS were relatively common
among LVSF articles (both around 15%) while ARDS comprised 25% of RVF articles. Transthoracic echocardiography
was the main echocardiography mode, in 87% of the articles for AET topic, followed by 81% for FM, 78% for LVDF, 70%
for LVSF and 63% for RVF. The percentage of items per study as well as the fraction of study reporting an item was low
or very low, except for FM. As an illustration, the left ventricular size was only reported by 56% of studies in the LVSF
topic, and half studies assessing RVF reported data on pulmonary artery systolic pressure.
Conclusion:  This analysis confirmed sub-optimal reporting of several items listed by an expert panel. The analysis will
help the experts in the development of guidelines for CCE study design and reporting.
Keywords:  Guidelines, Recommendations, Intensive care, Left ventricle, Right ventricle, Fluid management

Background
There is growing use of basic and advanced critical care
*Correspondence: antoine.vieillard‑[email protected] echocardiography (CCE) as a diagnostic and sequen-
10
Intensive Care Medicine Unit, Assistance Publique‑Hôpitaux de Paris,
tial monitoring tool for decision-making by intensive
University Hospital Ambroise Paré, 92100 Boulogne‑Billancourt, France
Full list of author information is available at the end of the article care physicians. The use of CCE has been defined as

© The Author(s) 2020. This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing,
adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and
the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material
in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material
is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the
permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat​iveco​
mmons​.org/licen​ses/by/4.0/.
Huang et al. Ann. Intensive Care (2020) 10:49 Page 2 of 13

echocardiography performed in critically ill patients function (LVDF); (4) fluid management (FM), and
by intensivists who also interpret the scan results [1], (5) advanced echocardiography techniques (AET,
although several CCE studies have involved cardiologists including speckle tracking and/or 3-D echocardiog-
or sonographers. This has been an area of rapid growth raphy studies only);
over the last decade with consequent demand for training c. the necessity to preventively establish a list of items
and accreditation processes, in addition to supporting that should be evaluated during the appraisal of the
evidence in the field [2, 3]. findings of the systematic search (see “Items and data
The Echocardiography Working Group of the Cardio- extraction”).
vascular Dynamics section of the European Society of d. the fact that the PRICES did not aim to create unrea-
Intensive Care Medicine (ESICM) recognizes that with a sonable standards of reporting CCE research which
growing CCE literature and huge heterogeneity in studies may bias against the publication of future important
identified by several systematic reviews and meta-analy- studies, but to give to the researchers a large amount
ses of CCE [4–10], there is a need to provide structured of information helping them in designing, conduct-
guidance for future CCE research methodology, report- ing and reporting their studies.
ing, and interpretation. The aim is to improve CCE
research data reporting for future research, to ultimately
support clinical decision-making in the monitoring, diag- Systematic review
nosis and treatment of critically ill patients. Literature search
The Echocardiography Working Group decided to per- The protocol of the systematic review was registered on
form first a comprehensive critical appraisal of the avail- PROSPERO database (CRD42018094450) on 1st May
able CCE literature to describe current reporting in order 2018. Literature searches using Medline and Embase
to provide evidence for the ultimate aim of PRICES (Pre- were made by SH (systematic review coordinator) and
ferred Reporting Items for Critical-care Echocardiography TP (professional librarian) in May 2018 and performed
Studies) recommendations. Here we report the results of separately for each topic/area with tailored search strate-
the systematic review describing the frequency of report- gies (see Additional file 1). The inclusion period was from
ing of items of possible importance for CCE research. 1st January 2000 to 31st December 2017. This period was
arbitrarily decided to produce an acceptable workload
Methods and because a large increase in the number of CCE publi-
Assembly of expert panel cations started since 2000 [3].
The PRICES project was initiated by the Echocardiog-
raphy Working Group of the ESICM. A total of 19 phy- Screening and studies appraisal
sicians with recognized expertise in the field of CCE Screenings were performed separately by experts for
were involved from different parts of the World (Europe each topic under the oversight of a designated team
n = 15, Oceania n = 3, North America n = 1). The first leader. Two experts screened each abstract retrieved
internal discussion regarding the PRICES project started from the search, and those satisfying all the following
in Vienna (September 25th and 26th, 2017). The authors criteria were included: (a) critical care population, (b)
requested and obtained endorsement by the ESICM. adult population, (c) reporting echocardiography data in
After extensive electronic correspondence, the experts’ the study, (d) clinical study, (e) English language, and (f )
group was first assembled in Brussels (March 17th, 2018) research articles with original data. A third expert was
where they agreed on: involved to resolve cases of disagreement. We excluded
studies where outcome from cardiac surgical conditions
a. the importance of supporting PRICES recommen- and techniques was the primary aim, and where patients
dations with a systematic review on the available were supported by extracorporeal membrane oxygena-
research that includes CCE data. This decision was tion or ventricular assist devices. The full-text articles of
made with the aim of providing a basis for a precise included abstracts were downloaded and were appraised
and critical appraisal of the utility of the reported in detail by two experts to ensure inclusion suitability.
information in current CCE literature according to Risk of bias assessment was beyond the scope of this
different domains (i.e. design, methodology, statis- appraisal and thus not performed.
tics, results reporting, etc.);
b. the need to split CCE literature according to specific Items and data extraction
areas (or “topics”) of interest in CCE research: (1) left Each included article was searched for a list of pre-deter-
ventricular systolic function (LVSF); (2) right ven- mined echocardiographic information (“preferred items”
tricular function (RVF); (3) left ventricular diastolic or simply “items”), the absence of which was deemed to
 Huang et al. Ann. Intensive Care (2020) 10:49 Page 3 of 13

potentially introduce bias in measurement, misinterpreta- Results

tion or non-reproducibility of the study results. Such items Figure  1 shows the flow diagram for the literature
were proposed during the first expert assembly and clas- search process. Medline and Embase returned 438
sified into “common” ones (study characteristics; patient and 157 articles of which 72 were duplicates. After the
characteristics; echocardiography information and pur- exclusion of 294 articles based on abstract screening,
pose; clinical information during echocardiography pro- 229 articles remained. Fifty-four articles were cross-
cedure; measurement reliability; statistical analysis) and referred during screening to other groups resulting in
“topic-specific” (Table 1). a total of 283 articles. The full-texts were appraised in
Most items were categorical and related to whether the detail, resulting in further exclusion of 74 articles. A
items had been reported or not, or in some cases how cer- total of 209 articles were finally included, some of which
tain information was collated. Double-data entry method were assigned to more than one topic group (LVSF 97,
(two different experts blinded to each other) was used for RVF 48, LVDF 51, FM 36, and AET 24) (Fig. 2a).
data extraction via a web-based database (REDCap hosted
at University of Sydney—https​://redca​p.sydne​y.edu.au).
Summary of reporting of “common items” (43 items)
Any discrepancy was resolved by a third expert of the same
A total of 43 items common to all CCE topics were
group (“adjudicator”), or eventually referring to a “grand
extracted. The values of FSi for each item are provided
adjudicator” for a final decision. The quality of data extrac-
according to the topic of interest for the main ones
tion was validated by an independent expert methodologist
(Figs. 3, 4, 5, 6, 7) and extensively as Additional files 2,
(GG). Briefly, a total of 20 articles were selected randomly
3, 4, 5, 6.
(proportionally to the total amount for each topic) and data
extracted was compared to those obtained by the experts.
A total of 11 discrepancies were found and, considering Study characteristics (3 items)
an average of ~ 60 items per study, the “error” rate was far All studies reported the sample size. Most studies were
below 1% per study. prospective observational (87%), while interventional
studies accounted for about 10% and the remaining
Data analysis were retrospective or post hoc studies.
Data analysis was conducted separately for each topic of
CCE interest. From the beginning it was clear that each Patients characteristics (12 items)
item did not carry the same importance in different areas Clinical context, age and gender had high FSi. The clini-
of CCE interest. The potential importance of each item and cal context varied among the CCE topics, with sepsis
recommendation for its reporting will be the object of the accounting for 40% to 54% in all topics except for FM
PRICES recommendation paper and are not discussed here where only 28% were sepsis-related and most were on
since this is a systematic descriptive non-clinical review. shock (44%). Shock and acute respiratory distress syn-
In the present study, data on item reporting appraisal are drome were relatively common among LVSF articles
summarized as percentage of items reported per study (both around 15%), while acute respiratory distress
(PIPS) and as fraction of studies reporting an item (FSi). syndrome comprised 25% of RVF articles (Fig.  2b).
PIPS was calculated as a percentage obtained from the Age and gender were reported in over 90% of studies
sum of items reported in a study divided by the total num- across all topics, but < 50% of studies reported height
ber of items: and weight, or body mass index. Among past medical
number of items reported in a study history data, atrial fibrillation was mentioned in about
PIPS =
total number of items
× 100%. 40% of studies, mostly as exclusion criteria. The rate
of reporting for other patients comorbidities was rela-
A low PIPS score means the study failed to report a sub- tively low (< 30%).
stantial number of items.
FSi was calculated as the total number of studies report-
Echocardiography: information and purpose (6 items)
ing a particular item divided by the total number of studies:
Transthoracic echocardiography was the main echo-
number of studies reporting an item cardiography mode: the highest was the AET topic
FSi = . (87%), followed by FM (81%), LVDF (78%), LVSF (70%)
Total number of studies included
and RVF (63%). Only 10–20% of the studies used
FSi can be viewed as the “popularity” of an item—the transesophageal echocardiography or both in each
higher FSi means the more studies reported it. The FSi topic. Apart from FM studies, the reports of image
was calculated for all the items. acquisition information were sub-optimal (e.g. < 40%
Huang et al. Ann. Intensive Care (2020) 10:49 Page 4 of 13

Table 1  Lists of various domains and preferred items

Domains and items

Common to all topics Study information (n = 3)

Study type, study design, sample size
Patients characteristics (n = 12)
Context
Age, gender, height and weight (or BMI)
History of hypertension, HFpEF, HFrEF, ischemic heart disease, atrial fibrillation, COPD, chronic renal failure, presence of pace‑
maker
Echocardiography information (n = 6)
Type of echocardiography; were data collected at end-expiration? Number of beats for data averaging? Was airway pressure
trace displayed on screen?
Vendor of ultrasound machine and software version
Clinical information at the time of echocardiography (n = 10)
Mode of ventilation; if mechanically ventilated tidal volume, plateau pressure and positive end-expiratory pressure
Cardiac rhythm, heart rate, blood pressure; inotropes, vasopressors and their doses
Measurement reliability (n = 8)
Feasibility; intra-observer and inter-observer variability; was observer blinded to treatment?
Echocardiographer professional training and experience in echocardiography
Reviewer’s professional training and experience in echocardiography
Statistics reporting (n = 4)
Was sample size and power calculation provided? Was analysis blinded? Were confounders addressed? Was internal validation
provided?
Topic-specific items LV systolic function (n = 15)
LV size, LV ejection fraction, LV fractional area change, Tissue Doppler Sʹ velocity, MAPSE, LV dP/dt, LV Tei index, LV strain or strain
rate, regional wall motion score
Cardiac output, stroke volume, presence of heart valve disease; patent foramen ovale; pericardial effusion, tamponade
RV function (n = 18)
RV end-diastolic diameter; RV end-diastolic area; RV-to-LV end-diastolic area ratio; TAPSE; RV fractional area change; tissue Dop‑
pler Sʹ velocity; RV Tei index; RV strain or strain rate; subjective rating of RV function; PAPs or TR peak velocity; PAAT​
Patent foramen ovale; pericardial effusion; tamponade; RV wall thickness; paradoxical septal motion; IAS bowing; IVC diameter
LV diastolic function (n = 15)
E/A ratio; tissue Doppler Eʹ velocity; E/Eʹ ratio; PAPs or TR peak velocity; mitral E propagation velocity; mitral E deceleration time;
pulmonary venous flow; left atrial size
Systolic, diastolic and mean blood pressure; chronic medications; criteria used for grading diastolic function; guidelines or refer‑
ence for criteria cited; technical details of measurements
Fluid management (n = 7)
Parameter used to predict FR, echocardiographic parameter to assess FR-to-volume challenge or passive leg raising
Was fluid responsiveness defined? Were technical details of measurements provided? Was reference (“gold”) standard for com‑
parison stated? Was description of the reference standard provided? Was echocardiography used as reference standard?
Advanced echocardiography techniques (n = 17)
Types of strain used in LV study; strain or strain rate used in LV study; myocardial layer analysed for LV strain study; RV longitudi‑
nal strain, RV longitudinal strain rate; number of cycles used in analysis; start time in cardiac cycle used in analysis, frame rate;
number of planes used in analysis; method of image exclusion, method of segments exclusion; details of image optimization
method; drift correction used
Number of beats used in 3-D analysis; frame or volume rate used in 3-D analysis; timing of respiratory cycle in 3-D analysis;
reference method in 3-D analysis
Items are divided in common to all critical care echocardiography studies and those of particular interest in a specific topic
BMI body mass index, COPD chronic obstructive pulmonary disease, FR fluid responsiveness, HFpEF heart failure with preserved ejection fraction, HFrEF heart failure
with reduced ejection fraction, IAS inter-atrial septum, IVC inferior vena cava, LV left ventricle, MAPSE mitral annulus plan systolic excursion, PAPs pulmonary artery
systolic pressure, PAAT​pulmonary artery acceleration time, RV right ventricle, TAPSE tricuspid annular plan systolic excursion, TR tricuspid regurgitation
 Huang et al. Ann. Intensive Care (2020) 10:49 Page 5 of 13

Fig. 1  Flowchart of the literature search. AET advanced echocardiography techniques, FM fluid management, LVDF left ventricular diastolic
function, LVSF left ventricular systolic function, RVF right ventricular function
Huang et al. Ann. Intensive Care (2020) 10:49 Page 6 of 13

Fig. 2  Number (a) and clinical context (b) of the included studies included into the systematic review, per topics. AET advanced techniques, ARDS
acute respiratory distress syndrome, ECMO extracorporeal membrane oxygenation, FM fluid management, LVDF left ventricular diastolic function,
LVSF left ventricular systolic function, RVF right ventricular function

Fig. 3  Radar plot of the fraction of studies reporting an item (FSi) in the left ventricular (LV) systolic function topic. HFrEF history of heart failure with
reduced ejection fraction, LVEF LV ejection fraction, LVFAC LV fractional area change, MAPSE mitral annulus plan systolic excursion, RWMAs regional
wall motion abnormalities, Sʹ maximal systolic velocity by tissue Doppler imaging at the mitral annulus. As example, an FSi score of 0.76 for LVEF
means that 76% of studies on LV systolic function reported LVEF
 Huang et al. Ann. Intensive Care (2020) 10:49 Page 7 of 13

Fig. 4  Radar plot of the fraction of studies reporting an item (FSi) in the right ventricular (RV) function topic. IVC inferior vena cava, LV left ventricle,
PAAT​pulmonary acceleration time, PAPs pulmonary artery systolic pressure, RVEDA RV end-diastolic area, RVEDD RV end-diastolic diameter, RV
FAC RV fractional area change, TAPSE tricuspid annulus systolic excursion, TR tricuspid regurgitation, Sʹ maximal systolic velocity by tissue Doppler
imaging at the tricuspid annulus. As example, an FSi score of 0.42 for RV-LV EDA ratio means that 42% of studies on RV function reported RV-LV EDA
ratio

reporting whether or not images were collected at exams, respectively. In most cases, critical care physicians
end-expiration, or number of cardiac cycles used for were responsible of both performing and reviewing the
averaging). exams. The rate of cardiologist involved in performing
echocardiography exams was 5% to 10% (LVSF, RVF and
Clinical information during echocardiography procedure (10 FM topic) and slightly higher for LVDF (16%) and AET
items) topic (37%). The involvement of cardiologist in review-
On average, over 65% articles in each topic reported the ing the exams were between 17 and 25%, except FM topic
heart rate and blood pressure, except for the FM topic were it was sensibly lower (6%). Sonographers were also
where > 80% of articles reported these information. Car- occasionally involved, but mainly in performing the stud-
diac rhythm was reported in almost 50% of the studies; ies only. The level of training of clinicians performing
the use of inotropes, vasopressors, and their doses were and reporting the exam was described in 41% and 25%
reported in 49%, 68% and 43%, respectively. Regarding of the articles, respectively. On average, 28% and 22% of
mechanical ventilation, the mode was described by 75% the studies reported intra-observer and inter-observer
of studies, while the ventilatory settings in the case of variabilities, respectively; 33% reported the feasibility of
mechanical ventilation, namely positive end-expiratory echocardiography.
pressure, plateau pressure and tidal volume, were rarely
reported (32%, 19% and 28%, respectively). Even in FM Statistics analysis (4 items)
group, only 50% to 60% of the studies reported this infor- Less than 25% of studies reported power and sample size
mation. Most studies (> 90%) did not report if airway calculation. The proportion of studies reporting if the sta-
pressures were displayed on the ultrasound monitor. tistical analyses were blinded varied grossly: 71% in AET,
43% in FM, 31% in LVDF, 27% both for LVSF and RVF.
Measurement reliability (8 items) Adjustment for confounders followed a similar trend.
Approximately 30% and 45% of the studies did not report
who performed and reviewed the echocardiography
Huang et al. Ann. Intensive Care (2020) 10:49 Page 8 of 13

Fig. 5  Radar plot of the fraction of studies reporting an item (FSi) in the left ventricular (LV) diastolic function topic. A atrial wave of transmitral
diastolic blood flow, BP blood pressure, E early wave of transmitral diastolic blood flow, Eʹ maximal diastolic early velocity by tissue Doppler imaging
at the mitral annulus, PAPs pulmonary artery systolic pressure, TR tricuspid regurgitation. As example, an FSi score of 0.59 for E/A ratio means that
59% of studies on LV diastolic function reported E/A ratio

Summary of topic‑specific items RV function (18 items, Fig. 4, Additional file 3)

The overall results of the values of the FSi of each topic- The average PIPS was low for this topic (19.1%). For stud-
specific item are presented in radar plots (Figs.  3, 4, 5, ies reporting RV dimensions, 42% used the RV-to-LV
6, 7). The greater the area of the plot itself, the better is end-diastolic areas ratio, 33% the RV end-diastolic area,
the overall reporting for topic-specific items in studies and 21% the RV end-diastolic diameter. 15% of stud-
regarding that topic. ies used subjective ratings of RV function, and 17% did
not report any parameter of function, except RV-to-LV
LV systolic function (15 items, Fig. 3, Additional file 2) end-diastolic areas ratio and paradoxical septal motion.
The average PIPS for studies included in the LVSF topic Half of studies reported pulmonary artery systolic pres-
was low (29.6%). LV ejection fraction was reported by sure (PAPs) directly or from tricuspid regurgitation jet
76% of studies, and in particular Simpson’s method, vis- velocity.
ual estimation or both were used in 54%, 20% and 2% of
the studies, respectively; 24% of studies did not indicate LV diastolic function (15 items, Fig. 5, Additional file 4)
their method for LV ejection fraction measurements. The average PIPS was 42.8%. E/Eʹ, Eʹ wave at mitral
For studies reporting LV size (56%), LV end-diastolic annulus on tissue Doppler imaging and E/A were more
diameter (23%), area (20%) and volume (28%) were used, commonly reported (67%, 63% and 58%, respectively) as
with some reporting more than one parameter (12%). compared with pulmonary artery pressure (PAPs, or sur-
For studies reporting Sʹ wave at mitral annulus on tissue rogates) and left atrial size (15% and 25%, respectively).
Doppler imaging (26%), 66% did not report the segments Regarding left atrial size, the parameter was reported as
used, while the remaining reported medial (septal) (11%), volume (14%), diameter (8%) and area (4%). PAPs meas-
lateral (14%) or average of the two walls (8%). ured directly was only reported in 4% of the studies, while
 Huang et al. Ann. Intensive Care (2020) 10:49 Page 9 of 13

Fig. 6  Radar plot of the fraction of studies reporting an item (FSi) in the fluid management topic. FR fluid responsiveness, PLR passive leg raising, VC
volume challenge. As example, an FSi score of 0.72 for FR definition means that 72% of studies on fluid management reported FR definition

11% of studies used tricuspid regurgitation jet velocity as commonly reported (42% and 46%, respectively). Only
surrogate for PAPs. Technical details of measurements 13% and 8% of studies reported circumferential and
were mostly reported (80%). The criteria used for evalu- radial strains, respectively. The type of LV strain used was
ating LVDF were quoted only in 69% of studies. not reported by 17% of studies. Acquisition and analysis
information were reported with a different degree, from
Fluid management (7 items, Fig. 6, Additional file 5) relatively high (frame rate 67%, number of planes used
The average PIPS was 78%. The methods used to assess for global strain 88%) to rather low (use of drift correc-
fluid responsiveness was reported by nearly all studies tion and segment exclusion 4%, clear image optimiza-
(97%), and various methods were used (volume chal- tion procedure 14%, no study reporting the start time of
lenge 72%; variations of stroke volume or its surrogates recording).
36%: change in inferior vena cava or superior vena cava, Regarding 3-D echocardiography, technical informa-
33% and 8%, respectively; passive leg raising 17%). Over tion were all seldomly reported.
90% of studies reported gave technical details of meas-
urements, but definition of fluid responsiveness was not Discussion
always clear (72%). Roughly three-quarters of studies This systematic review summarizes the research report-
reported if and which “gold” standard for comparison ing practice in CCE for studies published between year
was adopted to define fluid responders. 2000 and 2017. The aim of the systematic review was
to inspect past studies in order to describe reporting
attitude and to identify potential areas of weakness and
Advanced echocardiographic technique (17 items, Fig. 7, insufficient reporting, finally providing a robust evidence
Additional file 6) base for the expert panel to design recommendations
The average PIPS was 42%. A total of 13 items were iden- for standardized reporting of future studies. Our goal is
tified for speckle tracking studies and other four for the not to judge the quality of the past studies, nor to cre-
3-D studies. Most of ventricular strain studies were per- ate unreasonable standards that could limit in the future
formed on the LV (> 80%); strain was more used than the publication of interesting studies unable to report all
strain rate. Global and longitudinal strains were the most the necessary items. Of note, studies from authors of the
Huang et al. Ann. Intensive Care (2020) 10:49 Page 10 of 13

Fig. 7  Radar plot of the fraction of studies reporting an item (FSi) in the advanced echocardiography techniques (AET) topic. All parameters
but the last four in anticlockwise sense starting at 12 o’clock refers to strain echocardiography method. The last four refers to three-dimensional
echocardiography (3-D) method. LV: left ventricle, RV: right ventricle. As example, an FSi score of 0.82 for type of strain used for LV studies means
that 82% of studies reported the type of strain used to evaluate LV function

PRICES panel were evaluated in the same manner in this normal sinus rhythm. Another example, despite the fre-
systematic review, and we found many of them had the quent use of vasoactive drugs in intensive care which
same weaknesses and insufficiencies in reporting as the are known to affect the interpretation of most echocar-
other researchers. diographic variables, the presence and dosage of ino-
Our systematic review identified a considerable tropes and vasopressors were sub-optimally reported
heterogeneity between studies and between the dif- (49%, 68% and 43%, respectively). This would clearly
ferent fields of interest. For instance, studies in FM introduce a source of bias when comparing studies.
topic reported items in a higher number while those Furthermore, the mode of ventilation was described by
on LVDF topic lacked many items. Several items were three-quarters of studies; however, the values of posi-
under-reported despite their importance from either tive end-expiratory pressure, plateau pressure and tidal
a methodological or clinical perspective. A large vol- volume during the echocardiography examination were
ume of narrative information was collected during the only reported in a minority of cases despite ventilation
course of this work, but the discussion of all these find- settings are known to affect heart performance and
ings would make the manuscript unnecessarily long, so especially the RV function. Additionally, these omis-
we chose to present a limited sample to illustrate the sions will limit the validity of echocardiography param-
level of under-reporting of important items in CCE eters in the investigation of fluid responsiveness [14].
studies. For example—the presence of atrial fibrillation We also evaluated methodological aspects of echocar-
at the time of echocardiography was mentioned only in diography studies and data analysis in each study. Among
a minority of studies (mainly as exclusion criteria) while others, it appears that assessment for confounders, blind-
it is known that its incidence during critical illness is ing, identification of the person responsible of both per-
relatively high [11–13] and that it may induce cardiac forming and reviewing the echocardiography studies are
dysfunction (especially diastolic) and it complicates or far from being systematically reported. We also found
invalidates most echocardiographic measurements. under-reporting of the “topic-specific” items, where
Moreover, it precludes the use of AET which requires one ideally would expect higher reporting due to their
 Huang et al. Ann. Intensive Care (2020) 10:49 Page 11 of 13

specificity for the area of interest. For instance, the LV different items throughout the study period. We believe
ejection fraction was the most commonly used param- that changes over time in reporting certainly have hap-
eter to describe LVSF (76%), but information on LV size pened for certain items. The items of the AET are one
were provided in roughly half of studies. Information on clear example due to the novelty of this echocardiogra-
RV dimensions were under-reported to a similar extent phy modality, but also LVDF seems another field where
and RV wall thickness was seldom reported, despite the variations in reporting attitude have happened over the
role of these measurements in signalling the effect of time due to appearance of new guidelines [17] where the
chronic lung disease on the RV [15]. Surprisingly, in the use of some items has been reduced (i.e. deceleration
investigation of LVDF we found that in around one-third time and pulmonary venous flow) while it increased for
of cases the authors did not refer to existing guidelines others (tissue Doppler imaging, left atrial size and tricus-
[16, 17] and used their own criteria or quoted references pid regurgitation jet) [18].
other than guidelines. Similarly, in the study of the fluid Second, one can say that some results were quite
management over one-quarter of studies did not provide expected. In truth, we—as authors of CCE studies—
sufficient information about the reference (“gold”) stand- were somewhat surprised of the sub-optimal reporting
ard method used to assess fluid responsiveness. of items important for the interpretation of study find-
After reporting these examples, we would like to ings. In other words, we expected better performance
emphasize that the purpose of the present systematic in reporting from ourselves. This further highlights the
review is to provide solid evidence for the expert panel need for providing guidance in reporting CCE stud-
to design recommendations for the reporting of studies ies, even for people supposed to be experts in this field.
utilizing CCE, rather than to criticize the quality of the It is interesting for the researchers to note that in many
body of research or to create unreasonable standards. studies, the authors did not report parameters allowing
The information on the frequency of reporting will be accurate interpretation of study findings, such as the sub-
of course weighted against the importance of each item optimal reporting of LV size in studies regarding LVSF.
with the target of establishing the essential items that On the other side, the absence of reporting of certain
need mandatory reporting in CCE studies. The ulti- parameters are not surprising and as example we cannot
mate aim is to guide future CCE researchers to pursue a be surprised that dP/dt was rarely reported in studies on
standardized approach in study design and reporting to LVSF, although some intensivists suggested the useful-
enhance reproducibility and data homogeneity. This will ness of this parameter [19].
increase the external validity and the impact of individual Third, because we decided to perform our analysis by
studies, facilitating meaningful comparison and the pool- area of interest rather than by clinical situations which
ing of data in meta-analyses. Similar to the rationale for were regarded as too numerous and diverse, we acknowl-
the “PRISMA statement” [18], which provides structured edge that some items identified by the experts could be
guidance on the information that authors should report inappropriate or difficult in certain settings. The most
in systematic review and meta-analysis to improve data obvious situation is probably the use of CCE in cardiac
consistency and allowing meaningful pooling of results, arrest where nothing else than a qualitative evaluation
the next step of the PRICES project is to construct rec- is allowed, though it must be noted that studies on car-
ommendations based on this systematic review balanced diac arrest do not focus on the topics we selected for the
with expert opinion on the importance of the appraised appraisal.
items.
Conclusions
Limitations This systematic review critically appraised the reporting
Our study has some limitations. First, in chronologi- pattern in over 15 years of CCE literature, and represents
cal terms, our appraisal was limited to studies published the first step in PRICES, an ESICM endorsed project that
from 2000 until the end of 2017; although it is likely that will produce recommendations for the reporting of CCE
more recent studies have higher reporting scores, it is studies. This analysis confirmed sub-optimal reporting
also probable that articles published before 2000 had of a number of items, which if omitted are likely to bias
worse reporting scores. Therefore, we believe unlikely study interpretation and reproducibility of its results.
that FSi results would have changed significantly with the Despite all its limitations, the systematic description of
inclusion of more recent and older publications. Moreo- the reporting attitude in CCE studies will be helpful for
ver, it must be noted that we decided not to investigate the construction of PRICES recommendations.
the evolution of the frequency of the reporting of the
Huang et al. Ann. Intensive Care (2020) 10:49 Page 12 of 13

Supplementary information Nephrological, Anesthesiological and Geriatric Sciences, University of Rome,

“La Sapienza,” Policlinico Umberto Primo, Viale del Policlinico, Rome, Italy.
Supplementary information accompanies this paper at https​://doi. 14
 Bodleian Health Care Libraries, University of Oxford, Oxford, UK. 15 Medical
org/10.1186/s1361​3-020-00662​-y.
Intensive Care Unit, Amiens University Hospital, Amiens, France. 16 Department
of Intensive Care, Maastricht University Medical Centre+, University Maas‑
Additional file 1. Search strategies. tricht, Maastricht, The Netherlands. 17 Medical‑Surgical Intensive Care Unit,
Limoges University Hospital, Inserm CIC 1435, Limoges, France. 18 Division
Additional file 2. Summary of reporting of LVSF items. of Pulmonary, Critical Care and Sleep Medicine, Northwell Health LIJ/NSUH
Additional file 3. Summary of reporting of RVF items. Medical Center, Zucker School of Medicine, Hofstra/Northwell, Hempstead,
NY, USA.
Additional file 4. Summary of reporting of LVDF items.
Additional file 5. Summary of reporting of FM items Received: 29 October 2019 Accepted: 11 April 2020
Additional file 6. Summary of reporting of AET items.

Abbreviations
AET: Advanced echocardiography technique; CCE: Critical care echocardiogra‑ References
phy; ESICM: European Society of Intensive Care Medicine; FM: Fluid manage‑ 1. Mayo PH, Beaulieu Y, Doelken P, Feller-Kopman D, Harrod C, Kaplan A,
ment; FSi: Fraction of studies reporting an item; LVDF: Left ventricular diastolic et al. American College of Chest Physicians/La Societe de Reanimation de
function; LVSF: Left ventricular systolic function; PIPS: Percentage of items per Langue Francaise statement on competence in critical care ultrasonogra‑
study; PAPs: Pulmonary artery systolic pressure; RVF: Right ventricular function; phy. Chest. 2009;135(4):1050–60.
PRICES: Preferred reporting items for critical care echocardiography studies. 2. Vieillard-Baron A, Millington SJ, Sanfilippo F, Chew M, Diaz-Gomez J,
McLean A, et al. A decade of progress in critical care echocardiography: a
Acknowledgements narrative review. Intensive Care Med. 2019;45(6):770–88.
None. 3. Mayo P, Arntfield R, Balik M, Kory P, Mathis G, Schmidt G, et al. The ICM
research agenda on critical care ultrasonography. Intensive Care Med.
Authors’ contributions 2017;43(9):1257–69.
All authors participated to the systematic review. FS registered the systematic 4. Huang SJ, Nalos M, McLean AS. Is early ventricular dysfunction or dilata‑
review on Prospero database; SH and TP did the SR; all the authors screened tion associated with lower mortality rate in adult severe sepsis and septic
the abstracts and evaluated the studies for reporting; SH analysed the data; shock? A meta-analysis. Crit Care. 2013;17(3):R96.
FS and AVB checked the analysis. SH, FS and AVB wrote the initial manuscript. 5. Sanfilippo F, Corredor C, Arcadipane A, Landesberg G, Vieillard-Baron A,
All authors reviewed the manuscript. All authors read and approved the final Cecconi M, et al. Tissue Doppler assessment of diastolic function and rela‑
manuscript. tionship with mortality in critically ill septic patients: a systematic review
and meta-analysis. Br J Anaesth. 2017;119(4):583–94.
Funding 6. Sanfilippo F, Corredor C, Fletcher N, Landesberg G, Benedetto U,
This work was supported by the European Society of Intensive Care Medicine. Foex P, et al. Erratum to: diastolic dysfunction and mortality in septic
patients: a systematic review and meta-analysis. Intensive Care Med.
Availability of data and materials 2015;41(6):1178–9.
Yes. 7. Sanfilippo F, Corredor C, Fletcher N, Landesberg G, Benedetto U, Foex P,
et al. Diastolic dysfunction and mortality in septic patients: a systematic
Ethics approval and consent to participate review and meta-analysis. Intensive Care Med. 2015;41(6):1004–13.
Not applicable. 8. Sanfilippo F, Corredor C, Fletcher N, Tritapepe L, Lorini FL, Arcadipane A,
et al. Left ventricular systolic function evaluated by strain echocardiogra‑
Consent for publication phy and relationship with mortality in patients with severe sepsis or sep‑
All the authors gave their consent to co-sign the manuscript in its present tic shock: a systematic review and meta-analysis. Crit Care. 2018;22(1):183.
form. 9. Sanfilippo F, Johnson C, Bellavia D, Morsolini M, Romano G, Santonocito
C, et al. Mitral regurgitation grading in the operating room: a systematic
Competing interests review and meta-analysis comparing preoperative and intraopera‑
The authors declare that they have no competing interests. tive assessments during cardiac surgery. J Cardiothorac Vasc Anesth.
2017;31(5):1681–91.
Author details 10. Maybauer MO, Asmussen S, Platts DG, Fraser JF, Sanfilippo F, Maybauer
1
 Intensive Care Unit, Nepean Hospital, The University of Sydney, Sydney, DM. Transesophageal echocardiography in the management of burn
Australia. 2 Department of Anesthesia and Intensive Care, Policlinico-Vittorio patients. Burns. 2014;40(4):630–5.
Emanuele University Hospital, Catania, Italy. 3 Department of Intensive Care, 11. McIntyre WF, Um KJ, Cheung CC, Belley-Cote EP, Dingwall O, Devereaux
Erasme University Hospital, Univeristé Libre de Bruxelles, Brussels, Belgium. PJ, et al. Atrial fibrillation detected initially during acute medical illness: a
4
 Department of Anaesthesiology and Intensive Care, 1st Faculty of Medicine, systematic review. Eur Heart J Acute Cardiovasc Care. 2019;8(2):130–41.
Charles University and General University Hospital, Prague, Czech Republic. 12. Wetterslev M, Haase N, Hassager C, Belley-Cote EP, McIntyre WF, An Y,
5
 Department of Anaesthesiology and Intensive Care, Medical and Health et al. New-onset atrial fibrillation in adult critically ill patients: a scoping
Sciences, Linköping University, Linköping, Sweden. 6 Department of Anaes‑ review. Intensive Care Med. 2019;45(7):928–38.
thesiology and Critical Care Medicine, Vall d’Hebron University Hospital, 13. Kuipers S, Klein Klouwenberg PM, Cremer OL. Incidence, risk factors and
Barcelona, Spain. 7 Department of Perioperative Medicine, Bart’s Heart Centre outcomes of new-onset atrial fibrillation in patients with sepsis: a system‑
St. Bartholomew’s Hospital, W. Smithfield, London, UK. 8 CHIREC Hospitals, atic review. Crit Care. 2014;18(6):688.
Université Libre de Bruxelles, Brussels, Belgium. 9 Cardiothoracic Critical Care, 14. Vieillard-Baron A, Matthay M, Teboul JL, Bein T, Schultz M, Magder S, et al.
St Georges Hospital, St Georges University of London, London, UK. 10 Intensive Experts’ opinion on management of hemodynamics in ARDS patients:
Care Medicine Unit, Assistance Publique‑Hôpitaux de Paris, University Hospital focus on the effects of mechanical ventilation. Intensive Care Med.
Ambroise Paré, 92100 Boulogne‑Billancourt, France. 11 INSERM, UMR‑1018, 2016;42(5):739–49.
CESP, Team Kidney and Heart, University of Versailles Saint-Quentin en Yvelines, 15. Hilde JM, Skjorten I, Grotta OJ, Hansteen V, Melsom MN, Hisdal J, et al.
Villejuif, France. 12 Service de réanimation médicale, Hôpital Henri Mondor, Right ventricular dysfunction and remodeling in chronic obstructive
Assistance Publique-Hôpitaux de Paris, 51 Avenue du Maréchal de Lattre de pulmonary disease without pulmonary hypertension. J Am Coll Cardiol.
Tassigny, 94000 Créteil, France. 13 Department of Cardiovascular, Respiratory, 2013;62(12):1103–11.
 Huang et al. Ann. Intensive Care (2020) 10:49 Page 13 of 13

16. Nagueh SF, Appleton CP, Gillebert TC, Marino PN, Oh JK, Smiseth OA, et al. 19. Scolletta S, Bodson L, Donadello K, Taccone FS, Devigili A, Vincent JL, et al.
Recommendations for the evaluation of left ventricular diastolic function Assessment of left ventricular function by pulse wave analysis in critically
by echocardiography. Eur J Echocardiogr. 2009;10(2):165–93. ill patients. Intensive Care Med. 2013;39(6):1025–33.
17. Nagueh SF, Smiseth OA, Appleton CP, Byrd BF, 3rd, Dokainish H, Edvard‑
sen T, et al. Recommendations for the evaluation of left ventricular
diastolic function by echocardiography: an update from the american Publisher’s Note
society of echocardiography and the european association of cardiovas‑ Springer Nature remains neutral with regard to jurisdictional claims in pub‑
cular imaging. Eur Heart J Cardiovasc Imaging. 2016. lished maps and institutional affiliations.
18. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gotzsche PC, Ioannidis JP, et al.
The PRISMA statement for reporting systematic reviews and meta-anal‑
yses of studies that evaluate health care interventions: explanation and
elaboration. J Clin Epidemiol. 2009;62(10):e1–34.